简单的文本分类，用python实现了朴素贝叶斯和SVM
Simple text classification, realized with python Naive Bayes and SVM (2016-07-26, Unix_Linux, 4KB, 下载6次)
针对基于贝叶斯原理的序贯蒙特卡罗粒子滤波器出现退化现象的原因, 以无敏粒子滤波(U PF)、辅助粒子滤波
(A S IR) 及采样重要再采样(S IR) 等改进的粒子滤波算法为例, 对消除该缺陷的关键技术(优化重要密度函数及再采样) 进行了
分析研究。说明通过提高重要密度函数的似然度、引进当前测量值、预增和复制大权值粒子等方式, 可以有效改善算法性能。
最后通过对一无源探测定位问题进行仿真, 验证了运用该关键技术后, 算法的收敛精度和鲁棒性得到进一步增强。
Abstract:W e analyze the degeneracy phenomenon of sequen t ialMon te Carlo part icle f ilters based on
bayesian theo rem , pu t focu s on the key techn iques ( good cho ice of impo rtance den sity and u se of
resamp ling ) to reduce it s effect s. Several imp roving schemes such as the U n scen ted Part icle F ilters
(U PF) , the A ux iliary Samp ling Impo rtance Resamp ling (A S IR ) and the Samp ling Impo rtance Resamp ling
(S IR ) algo rithm s are in t roduced to illu st rate th rough increasing the likelihood of the impo rtance den sity o r
inco rpo rat ing new measu remen t, o r rep licat ing part icles w ith large w eigh t s w ith in the generic f rame of
part icle f ilters, the convergence accu racy and robu stness behavio rs of the algo rithm can be effect ively
imp roved. A typ ical passive detect ion and locat ion p rob lem is simu lated to p rove above conclu sion s. (2014-01-21, Unix_Linux, 291KB, 下载24次)
rtsp服务程序，客服端可用VLC播放器登陆
rtsp service program, client service available VLC player landing (2013-11-18, Unix_Linux, 355KB, 下载11次)
《集体智慧编程》(programming collective intelligence building smart web 2.0 applications)以机器学习与计算统计为主题背景，专门讲述如何挖掘和分析web上的数据和资源，如何分析用户体验、市场营销、个人品味等诸多信息，并得出有用的结论，通过复杂的算法来从web网站获取、收集并分析用户的数据和反馈信息，以便创造新的用户价值和商业价值。全书内容翔实，包括协作过滤技术（实现关联产品推荐功能）、集群数据分析（在大规模数据集中发掘相似的数据子集）、搜索引擎核心技术（爬虫、索引、查询引擎、pagerank算法等）、搜索海量信息并进行分析统计得出结论的优化算法、贝叶斯过滤技术（垃圾邮件过滤、文本过滤）、用决策树技术实现预测和决策建模功能、社交网络的信息匹配技术、机器学习和人工智能应用等。
The collective wisdom of programming " (programming collective intelligence building smart web 2.0 applications) Machine Learning and Computational Statistics background data and resources devoted to mining and analysis on the web, how to analyze the user experience, marketing, personal tastes, and many other information and draw useful conclusions, through a complex algorithm obtained from the web site, the collection and analysis of user data and feedback information, in order to create a new user value and commercial value. The book is informative, including collaborative filtering technology (associated product recommendation function), cluster analysis (in large-scale data sets to explore similar data subset), the core technology of the search engine (reptiles, index, query engine, pagerank algorithm), search massive amounts of information and statistical analysis to draw conclusions optimization algorithm, Bayesian filtering technology (spam filtering, text filtering), decisi (2012-08-21, Unix_Linux, 27743KB, 下载7次)
最优化算法，应用有限内存拟牛顿方法（Limited Memory (variable-storage)quasi-Newton method）求解高维最优化问题，使用更多的内存将使算法更有效。
optimization algorithm, Application limited memory quasi-Newton method (Limited Memory (variable-stora ge) quasi--Newton method) for high-dimensional optimization problem, use more memory will enable more effective algorithm. (2006-12-01, Unix_Linux, 29KB, 下载142次)
dbacl是一个通用目的的digramic贝叶斯文本分类器。它可以学习你提供的文本，接着用 学到的类别来比较新的输入。它可以用于垃圾邮件过滤，或在你自己的外壳脚本中使用。
dbacl is a general-purpose text digramic Bayesian classifier. It can learn to provide you with the text, then uses learned relatively new type of input. It can be used in spam filtering, or in your own shell scripts to use. (2006-05-22, Unix_Linux, 348KB, 下载51次)